library(tidyverse)
library(plotly)
library(gganimate)
library(gifski)
data_cleaned <- read.csv("../data/cleaned.csv")

rate_year <- data_cleaned %>%
        group_by(year) %>%
        summarize(rate = length(which(college_attendance_status=="Yes"))/n())

pyear = ggplot(rate_year, aes(x=year, y=rate)) +
        geom_line() +
        geom_point() +
        labs(title = "Trend Over Time", x="Year", y= "College Attendence Rate") +
        scale_x_continuous(breaks = seq(2008, 2018, 2)) +
        transition_reveal(year)
animate(pyear, duration = 5, fps = 20, width = 400, height = 400, renderer = gifski_renderer())

rate_city <- data_cleaned %>%
        group_by(year, metropolitan_status) %>%
        summarize(rate = length(which(college_attendance_status=="Yes"))/n())
        # pivot_wider(names_from =metropolitan_status, values_from = rate)

pcity = ggplot(rate_city, aes(x=year, y=rate, color = metropolitan_status)) +
        geom_line() +
        geom_point() +
        scale_x_continuous(breaks = seq(2008, 2018, 2)) + 
        labs(title = "Trend Over Time by Metropolitan", x="Year", y= "College Attendence Rate") +
        transition_reveal(year)
animate(pcity, duration = 5, fps = 20, width = 600, height = 400, renderer = gifski_renderer())

# plot_ly(rate_city, x=~year, y=~`in metropolitan and central city`, name = "in metropolitan and central city", type = "scatter", mode = "line") %>%
#         add_trace(y=~`in metropolitian but mixed of central city`, name = "in metropolitian but mixed of central city", type = "scatter", mode = "line") %>%
#         add_trace(y=~`in metropolitian but not in central city`, name = "in metropolitian but not in central city", type = "scatter", mode = "line") %>%
#         add_trace(y=~`mixed of metropolitian status`, name = "mixed of metropolitian status", type = "scatter", mode = "line") %>%
#         add_trace(y=~`Not in metropolitan area`, name = "Not in metropolitan area", type = "scatter", mode = "line")# plot_ly(rate_city, x=~year, y=~`in metropolitan and central city`, name = "in metropolitan and central city", type = "scatter", mode = "line") %>%
#         add_trace(y=~`in metropolitian but mixed of central city`, name = "in metropolitian but mixed of central city", type = "scatter", mode = "line") %>%
#         add_trace(y=~`in metropolitian but not in central city`, name = "in metropolitian but not in central city", type = "scatter", mode = "line") %>%
#         add_trace(y=~`mixed of metropolitian status`, name = "mixed of metropolitian status", type = "scatter", mode = "line") %>%
#         add_trace(y=~`Not in metropolitan area`, name = "Not in metropolitan area", type = "scatter", mode = "line")

rate_state <- data_cleaned %>%
        group_by(year, states) %>%
        summarize(rate = length(which(college_attendance_status=="Yes"))/n()) %>%
        mutate(rank = min_rank(-rate)) %>%
        ungroup()
        # pivot_wider(names_from =metropolitan_status, values_from = rate)

pstate <-ggplot(rate_state,aes(rank,group=states,fill=as.factor(states),color=as.factor(states))) +
 geom_tile(aes(y = rate/2,height = rate, width = 0.9), alpha = 0.8, color = NA) +
 geom_text(aes(y = 0, label = paste(states, ' ')), vjust = 0.2, hjust = 1) +
 geom_text(aes(y=rate,label = paste(' ',rate)), hjust=0)+
 coord_flip(clip = 'off', expand = TRUE) +
 scale_y_continuous(labels = scales::comma) +
 scale_x_reverse() +
 guides(color = FALSE, fill = FALSE) +
 theme_minimal() +
 theme(
 plot.title=element_text(size=25, hjust=0.5, face='bold', colour='grey', vjust=-1),
 plot.subtitle=element_text(size=18, hjust=0.5, face='italic', color='grey'),
 plot.caption =element_text(size=8, hjust=0.5, face='italic', color='grey'),
 axis.ticks.y = element_blank(), 
 axis.text.y = element_blank(), 
 plot.margin = margin(1,1,1,4, 'cm')
 )


pstate <- pstate + transition_states(states = year, transition_length = 4, state_length = 1) + 
 ease_aes('cubic-in-out') +
 labs(title = 'College Attendance Rate per Year : {closest_state}', 
 x='',y='Total Suicides per year')

animate(pstate,fps = 20,duration = 30, width = 950, height = 750, renderer = gifski_renderer())

# 
# pstate = ggplot(rate_state, aes(x=rank, y=rate, color = states)) +
#         geom_tile(aes(y = rank, height = rate, width = .9, color = NA))+
#         geom_text(aes(y = 0, label = paste(states, ' ')), vjust = 0.2, hjust = 1) +
#  geom_text(aes(y=rate,label = paste(' ',rate)), hjust=0)+
#         coord_flip(clip = 'off', expand = TRUE) +
#         scale_y_continuous(labels = scales::comma) + 
#         scale_x_reverse() +
#         guides(color = FALSE, fill = FALSE) +
#         labs(title = "College Attendance Rate by State", y="Year", x= "College Attendence Rate")
#         transition_reveal(year)
#         
# pstate

data_omitna <- na.omit(data_cleaned)
rate_income <-  data_omitna%>%
                mutate(income = car::recode(`anually_family_income`, "lo:20000 = 'low';94500:hi='high';else = 'middle'")) %>%
        # mutate(income = car::recode(`anually_family_income`, "lo:median(data_omitna$anually_family_income)*0.3 = 'low';median(data_omitna$anually_family_income)*2:hi='high';else = 'middle'")) %>%
        group_by(year, income) %>%
        summarize(rate = length(which(college_attendance_status=="Yes"))/n())
        # pivot_wider(names_from =income, values_from = rate)

pincome = ggplot(rate_income, aes(x=year, y=rate, color = income)) +
        geom_line() +
        geom_point() +
        scale_x_continuous(breaks = seq(2008, 2018, 2)) + 
        labs(title = "Trend Over Time by Income", x="Year", y= "College Attendence Rate") +
        transition_reveal(year)
animate(pincome, duration = 5, fps = 20, width = 600, height = 400, renderer = gifski_renderer())

# plot_ly(rate_income, x=~year, y=~`high`, name = "high income", type = "scatter", mode = "line") %>%
#         add_trace(y=~`middle`, name = "middle income", type = "scatter", mode = "line") %>%
#         add_trace(y=~`low`, name = "low income", type = "scatter", mode = "line")

rate_race <- data_cleaned %>%
        group_by(year, race) %>%
        summarize(rate = length(which(college_attendance_status=="Yes"))/n())
        # pivot_wider(names_from =race, values_from = rate)

prace = ggplot(rate_race, aes(x=year, y=rate, color = race)) +
        geom_line() +
        geom_point() +
        scale_x_continuous(breaks = seq(2008, 2018, 2)) + 
        labs(title = "Trend Over Time by Race", x="Year", y= "College Attendence Rate") +
        transition_reveal(year)
animate(prace, duration = 5, fps = 20, width = 600, height = 400, renderer = gifski_renderer())

# plot_ly(rate_race, x=~year, y=~`White`, name = "White", type = "scatter", mode = "line") %>%
#         add_trace(y=~`African American`, name = "African American", type = "scatter", mode = "line") %>%
#         add_trace(y=~`American Indian or Alaska Native`, name = "American Indian or Alaska Native", type = "scatter", mode = "line") %>%
#         add_trace(y=~`Other Asian or Pacific Islander`, name = "Other Asian or Pacific Islander", type = "scatter", mode = "line") %>%
#         add_trace(y=~`Chinese`, name = "Chinese", type = "scatter", mode = "line") %>%
#         add_trace(y=~`Japanese`, name = "Japanese", type = "scatter", mode = "line") %>%
#         add_trace(y=~`Others`, name = "others", type = "scatter", mode = "line")

 rate_noc <-  data_omitna%>%
                mutate(noc = car::recode(`number_of_children`, "0 = '0'; 1:2 = '1-2';3:hi='3+'")) %>%
        group_by(year, noc) %>%
        summarize(rate = length(which(college_attendance_status=="Yes"))/n())
        # pivot_wider(names_from =noc, values_from = rate)

pnoc = ggplot(rate_noc, aes(x=year, y=rate, color = noc)) +
        geom_line() +
        geom_point() +
        scale_x_continuous(breaks = seq(2008, 2018, 2)) + 
        labs(title = "Trend Over Time by Number of Children", x="Year", y= "College Attendence Rate") +
        transition_reveal(year)
animate(pnoc, duration = 5, fps = 20, width = 600, height = 400, renderer = gifski_renderer())

# plot_ly(rate_noc, x=~year, y=~`0`, name = "0 children", type = "scatter", mode = "line") %>%
#         add_trace(y=~`1-2`, name = "1-2 children", type = "scatter", mode = "line") %>%
#         add_trace(y=~`3+`, name = "3+ children", type = "scatter", mode = "line")
# rate_noc$frame = rate_noc$year
# plot_ly(rate_noc, x = ~year, y = ~rate, split = ~noc, frame = ~frame, type = 'scatter', mode = 'lines', line = list(simplyfy = F)) 
        # animation_opts(frame = 100, transition = 0, redraw = FALSE) %>%
        # animation_slider(hide = T) %>% 
        # animation_button(x = 1, xanchor = "right", y = 0, yanchor = "bottom")

n_school <- data_omitna %>%
        mutate(income = car::recode(`anually_family_income`, "lo:20000 = 'low';94500:hi='high';else = 'middle'"),
               income = factor(income, levels = c("low","middle","high")))
school_incomee = 
        n_school %>% 
        group_by(year, income) %>%
        summarize(income_group = n()) 
school_typee = 
        n_school %>% 
        group_by(year, income, school_type) %>% 
        summarize(number = n()) %>% 
        left_join(school_incomee, by = c("year","income")) %>% 
        mutate(percentage = 100*number/income_group)
plot_ly(school_typee, x=~year, y = ~percentage, color = ~school_type, frame = ~income, type = "scatter", mode = "line")

# plot_ly(rate_school, x=~year, y=~`Private school`, name = "Private school", type = "scatter", mode = "line") %>%
#         add_trace(y=~`Public school`, name = "Public school", type = "scatter", mode = "line") %>%
#         add_trace(y=~`Not enrolled`, name = "Not enrolled", type = "scatter", mode = "line")